Skip to Main content Skip to Navigation

Raisonnement distribué dans un environnement ambiant

Amina Jarraya 1, 2
1 ACMES-SAMOVAR - Algorithmes, Composants, Modèles Et Services pour l'informatique répartie
SAMOVAR - Services répartis, Architectures, MOdélisation, Validation, Administration des Réseaux
Abstract : Pervasive Computing and Ambient Intelligence aim to create a smart environment withnetworked electronic and computer devices such as sensors seamlessly integrating into everyday life and providing users with transparent access to services anywhere and anytime.To ensure this, a system needs to have a global knowledge of its environment, and inparticular about people and devices, their interests and their capabilities, and associated tasks and activities. All these information are related to the concept of context. This involves gathering the user contextual data to determine his/her current situation/activity; we also talk about situation/activity identification. Thus, the system must be sensitive to environment and context changes, in order to detect situations/activities and then to adapt dynamically.Recognizing a situation/an activity requires the definition of a whole process : perception of contextual data, analysis of these collected data and reasoning on them for the identification of situations/activities.We are particularly interested in aspects related to the distributed modeling of the ambient environment and to those related to distributed reasoning in the presence of imperfect data for the identification of situations/activities. Thus, the first contribution of the thesis concerns the perception part. We have proposed a new perception model that allows the gathering of raw data from sensors deployed in the environment and the generation of events.Next, the second contribution focuses on the observation and analysis of these events by segmenting them and extracting the most significant and relevant features. Finally, the last two contributions present two proposals concerning the distributed reasoning for the identification of situations/activities ; one represents the main contribution and the other represents its improved version overcoming certain limitations. From a technical point of view, all these proposals have been developed, validated and evaluated with several tools.
Complete list of metadatas

Cited literature [174 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Tuesday, November 12, 2019 - 5:08:23 PM
Last modification on : Sunday, September 27, 2020 - 3:24:55 AM
Long-term archiving on: : Thursday, February 13, 2020 - 6:33:49 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02360245, version 1


Amina Jarraya. Raisonnement distribué dans un environnement ambiant. Réseaux et télécommunications [cs.NI]. Université Paris Saclay (COmUE); Université de Tunis El-Manar. Faculté des Sciences de Tunis (Tunisie), 2019. Français. ⟨NNT : 2019SACLL011⟩. ⟨tel-02360245⟩



Record views


Files downloads